of The American Society for Information Science

Vol. 26, No. 3

February/March 2000

Go to
Bulletin Index

bookstore2Go to the ASIS Bookstore

  Copies

Social Informatics and Aviation Technology

by  Terry von Thaden

To date, human factors research in aviation has focused primarily on interface and human-computer interaction problems. Through research, I've discovered that much of the time, information needed to avoid an accident is present, yet it is not used effectively. My concern in this and other temporal, safety-critical areas is with the social and situated use of information, and its application in accident mitigation. I bring to this problem a background in aviation I am a pilot and have studied air traffic control and in human factors (engineering psychology), and I have studied both the overall organizational factors and the critical minutiae of performance factors that lead to aviation accidents.

The Flight Deck as Sociotechnical System

Until about 40 years ago, aviation accidents were primarily due to equipment failure or environmental factors. When the introduction of the turbine engine in the late 1950s made aircraft more reliable, the primary cause of accidents became human error. Today, aviation operations have become increasingly computer-mediated, forcing pilots to become agents in the transformation of data from digitized interfaces into meaningful information. It is from within this process the accumulation, internalization and dissemination of information that errors and mistakes persist.

Computer technology and digital displays have improved pilot access to information, but have not necessarily addressed the use of information, once displayed. A case in point is the persistence of Controlled Flight into Terrain (CFIT) accidents, in which an aircraft with no mechanical problem is flown, under control, into terrain without awareness on behalf of the crew. CFIT is the third-ranking cause of fatalities among air carriers in the United States. Much time, effort and money have been spent developing technologies and terrain displays to alleviate CFIT, yet it still happens. I believe that a social informatics approach to this problem would be valuable. By conducting situated studies to identify the true use of information on the flight deck, human factors specialists will begin to understand pilots' information needs and behavior in their actual environment.

It is estimated that human error contributes to 80% of all aviation accidents through such particulars as captain's authority, crew climate and decision skills. With advanced equipment to aid pilot flight duties has come an increase in workload and increased demands on working memory and attention allocation. Pilot duties include information assessment and integration to form a cohesive representation of the current and future state of the aircraft, known as situation awareness. Pilots require clearly defined information with minimal to no distraction, especially during periods of high workload, to achieve effective situation awareness. It has been suggested that the use of external alarms adds a layer of defense to the crew. Out of this comes the development of technological terrain warning systems such as Enhanced Ground Proximity Warning Systems (EGPWS), designed to aid the pilots in avoiding CFIT, and pick up where other terrain warning systems are perceived to have failed.

Technological advances may add to flight efficiency, but they may not aid the operators as they subsequently learn to rely on these systems and become de-skilled in their performance. The invisibility of flight processes through advanced automation takes the crew out of the information-seeking loop. The crew is taught to trust that automated systems are working properly and that there is no need to be aware of each automated change to the state of the aircraft. Over-reliance on technology may instill a false sense of safety in the flight crew and promote a complacent information-seeking environment in which the crew uses the technology to provide, rather than support, situation awareness. The impact of technology on the crew's ability to manage workload may yield benefits or impose new burdens leading to task escalation. New equipment will not deter CFIT without proper consideration of information use in the context of the competing demands facing pilots.

The Need for a Social Informatics Approach

Historically, approaches to understanding pilot needs and actions on the flight deck have been studied through the lens of individual cognition, performance and perception. While researchers in a range of fields have published numerous studies regarding communication networks and knowledge structures, it appears that there has yet to be an ethnographic study of the nature and use of information among aviation crews. Aviation system design will benefit from explicit study of both the social context in which flight crews work and actual information seeking and use on the flight deck. As of this writing, there has been no close social study of the effects of EGPWS technology on the information ecology to use Nardi and O'Day's terminology of the flight deck, yet government agencies are deliberating mandating its presence. My previous aviation experience and research suggest several social factors that may account, at least in part, for aviation accidents.

According to the distributed cognition theory developed by Edward Hutchins in his study of shipboard navigation (1996): "Social organizational factors often produce group properties that differ considerably from the properties of individuals." On the flight deck, then, information processed collaboratively by multiple crew members may be more robust than information processed by each individual, yet it requires social and organizational devices for continued support of group information retrieval in increasingly complex situations. Unfortunately, a damaging social factor in commercial aviation is the institutional flaw that allows pilots to individually "bid" for trips, a practice that does not maintain consistency in crew makeup over time. The flight deck's social atmosphere must be negotiated for every change in crew, since interaction patterns do not normally carry over between flight assignments. Thus, it is difficult to quickly negotiate common ground and lay a framework for joint activity. The facilitation of shared information interpretation is hampered by this lack of cultural and experiential common ground among flight crews. It is only through the situated activity of the flight that crew members negotiate the meaning and implications of information.

Diagnosing the Causes of Aviation Accidents

Many CFIT accidents have happened with aircraft already equipped with terrain warning devices. From the original National Transportation Safety Board database, I examined 15 CFIT accidents involving regional carriers. I provided experienced pilots with narrative information regarding these accidents and asked them to estimate the probability that pilots would have avoided the crashes had the aircraft been equipped with EGPWS. Based on the ratings, the overall median probability that these crashes would be avoided using EGPWS was 59%. While this type of finding bodes in favor of the new technology, the written survey comments and remarks made in follow-up interviews suggest something quite different. For example, below are several positive or neutral numerical ratings, paired with the explanation of those ratings given during the interviews:

    80%: "Well, you look at this and there is no crew coordination. It's a strong possibility that they would have gone missed approach with the information from EGPWS."

    80%:  "The crew was 'miles' behind the aircraft."

    60%:  "Poor cockpit crew coordination, poor operations, loss of situation awareness."

    50%: "These jokers are never in control of the approach. Would they have broken off with EGPWS, I doubt it."

The interview responses certainly do not convey a strong belief that EGPWS would be effective in preventing CFIT accidents. I draw two conclusions from this study that both point to the need for socially grounded studies of flight deck activities. One is that quantitative survey data are inadequate for full interpretation of perceived accident causes. The second is that a more complete knowledge of how errors occur is desperately needed.

Certainly understanding the limits on cognition and human information processes helps illuminate how digital flight information should be presented for use. These factors must be considered before decision support tools are designed and put into practice. But understanding the social situations in which errors occur may allow us to answer the question of why some categories of mistakes persist over time and others seemingly appear to get resolved. Cognitive, individual error causation is just part of the picture. It comprises one form of activity or orientation that is part of the more comprehensive ecology of information disasters on the flight deck. As soon as there is disagreement about what constitutes an error and how it was made as there clearly was in my study the door is open to ongoing debate concerned with how mistakes can be identified and rectified.

It is apparent that the causal factors of CFIT constitute a gray area that can be attributed, on the surface, to poor crew communication skills, flawed situation awareness and/or equipment failure or error. Social, technical and cultural factors must be considered when studying error and its amelioration on the flight deck. To determine the acceptability of a new technology and approve its implementation from studies conducted in laboratory settings seems presumptuous, if not downright dangerous. Requiring equipment with questionable effectiveness to remedy accidents whose main causes lie in the social use of information, rather than equipment use, has the potential to mask, rather than treat, the real factors that cause accidents. CFIT begins with a chain of events that manifests itself in the dynamic environment of the flight deck. Through careful consideration of the multitasking environment and the temporal nature of the information ecology of the flight deck, systems developers can begin to understand proper information processing and display.

Pilots often fail to detect or respond correctly to warnings. Equipment is not a replacement for poor communication, poor procedure or inadequate training. It is recommended that situated studies of the information ecology of the flight deck, specifically as it relates to the use of terrain warning equipment, be performed. Only then will it be understood exactly how the information from terrain warning devices is used, and only then can recommendations regarding the regulation of such devices be considered.

Further Reading

Hutchins, E. (1996). Cognition in the Wild. Cambridge, MA: MIT Press.

Terry von Thaden, a doctoral student in the Graduate School of Library and Information Science at the University of Illinois at Urbana-Champaign, can be reached by e-mail at vonthade@uiuc.edu


How to Order

@ 2000, American Society for Information Science